Physical activity and dietary practices are well-documented in the etiology of cardiovascular diseases (CVD). While it is suspected that the behavioral roots of obesity, hypertension, hyperlipidemia, and clinical CVD are present in childhood, there have been relatively few attempts to study the early development of dietary and physical activity behaviors. A three-year grant has been devoted to developing the methology and recruiting the cohort in preparation for a longitudinal study. Numerous new measures have been developed and evaluated. Continuation of the current project will add significantly to the growing knowledge base because half of the sample consists of Hispanic children and half are Anglo. The proposed project will study the determinants of dietary and physical activity behaviors in 350 4 year old Hispanic and Anglo children. The primary aims of the 5 year prospective observational study are as follows: (a) to describe the development and degree of tracking (relative stability over time) of physical activity and dietary intake of saturated fat, sodium, and calories in preschool children over a 5 year period; (b) to identify important determinants of physical activity and dietary habits in young children, with an emphasis on family determinants; (c) to describe the development and determinants of salt taste preference in young children; and (d) to determine relationships between dietary and physical activity habits in young children and physiological indicators of risk, such as blood pressure, body fat, and body mass. The preschoolers and their families will be measured every 6 months for 5 years over a critical period of development. Children are observed during free play periods and meals during school visits and evening home visits. Determinants were selected based on social learning theory, and include sociodemographic variables (e.g., ethnic group, education, acculturation of Hispanics); parental risk status; parental modeling, prompts and knowledge; teacher and peer prompts, salt taste preference of parents and children; child variables (e.g., ponderosity, dietary and physical activity knowledge and preferences); television influences; and opportunities for physical activity and food choice. The growth curve model will be used in the analysis of data. Results of the study will add significantly to the data base on health habits in Mexican, American and Anglo preschool children. Such data can guide priorities for intervention among this high risk group.